Discussion

The importance of NAA and its metabolism in CNS function is highlighted by the fatal outcome of Canavan disease in which ASPA activity is disrupted. Several physiological roles have been proposed for this CNS-unique substance (Tsai and Coyle, 1995), including neuronal molecular water pump (Baslow, 2000; 2003), neuronal osmoregulation (Taylor et al., 1995), intracellular anion pool (McIntosh and Cooper, 1965), a storage form of aspartate (Birken and Oldendorf, 1989), both precursor and metabolite of NAAG (Cangro et al., 1987), a storage form of acetate for acetyl-CoA formation (Mehta and Namboodiri, 1995), and source of acetyl groups for myelin lipid synthesis (Chakraborty et al., 2001). Biochemical studies showing ASPA occurrence predominantly in white matter of brain (McIntosh and Cooper, 1965; D'Adamo et al., 1973; Goldstein, 1976; Kaul et al., 1991) provided early suggestions of a role in myelin formation and/or turnover, consonant with developmental data in the rat showing maximal enzyme activity (Goldstein, 1976) and in situ hybridization (Kirmani et al., 2003) at the peak of myelination. Initially described as a soluble enzyme (D'Adamo et al., 1973; 1977), ASPA was also claimed to be membrane-bound as well (Goldstein, 1976), a result supported by our findings (Chakraborty et al., 2001). Kaul et al. (1991) pointed to the requirement of detergent for solubilization as evidence of membrane association. This may also explain our finding of an hour or so delay in biochemical detection of ASPA activity in myelin (Fig. 4); this was not observed with brain homogenate which included cytosolic- as well as membrane-bound activity (Chakraborty et al., 2001). A recent study in which biochemical analysis failed to detect ASPA in myelin employed only 1 hr of incubation (Madhavarao et al., 2004). Our quantitative biochemical analyses indicated similar specific activities of ASPA in myelin and brain cytosol. These likely refer to subfractions of OLs, the cell type in which ASPA is localized (Baslow et al., 1999).

[NAA] mil Hours of Incubation

Figure 4. Biochemical detection of ASPA in myelin. Left panel: variation of myelin ASPA activity with NAA concentration; inset = Lineweaver-Burk plot. Right panel: variation of ASPA activity with time. Myelin required one hr of contact with detergent before activity commenced; it achieved parity with brain homogenate at 4 hr. Reproduced from Chakraborty et al. (2001) with permission from Blackwell Publishing.

[NAA] mil Hours of Incubation

Figure 4. Biochemical detection of ASPA in myelin. Left panel: variation of myelin ASPA activity with NAA concentration; inset = Lineweaver-Burk plot. Right panel: variation of ASPA activity with time. Myelin required one hr of contact with detergent before activity commenced; it achieved parity with brain homogenate at 4 hr. Reproduced from Chakraborty et al. (2001) with permission from Blackwell Publishing.

Evidence supporting a role for myelin-associated ASPA in providing acetyl groups for myelin lipid synthesis was provided in the current study of myelin lipids from brains of

ASPA-null (KO) mice, previously created by Matalon et al. (2000). Absence of this enzyme in brain would preclude liberation and hence utilization of acetyl groups from NAA. The yield of isolated myelin from the brains of such animals was ~30% less than from normal brain, consistent with the myelin aberration found in Canavan disease brain (Matalon et al., 1995). In addition to that deficit we found ~20% reduction of phospholipid relative to protein in the KO myelin; preliminary studies suggest this may apply to specific phospholipids rather than phospholipids in general. Of particular interest was the significant reduction of cerebroside, pronounced deficit occurring in the subgroup of cerebrosides containing unsubstituted fatty acids. This group (Cereb 1, Fig. 1A) was depleted ~65% in myelin and somewhat more in whole brain, attributed to reduced myelination plus deficient synthesis of that lipid relative to myelin protein. The final step in cerebroside synthesis, catalyzed by galactosyltransferase, was previously shown to occur bimodally in myelin and microsomes (Neskovic et al., 1973; Costantino-Ceccarini and Suzuki, 1975; Koul and Jungalwala, 1986); our finding suggests the same may apply to synthesis of the ceramide unit of cerebrosides, the component that would utilize NAA-acetyl.

The above mentioned early studies showing incorporation of the acetyl group of NAA into brain and myelin lipids employed extracellular application of NAA in the form of intracerebral injection or tissue slice incubation, leaving open the question of intercellular transfer. A more recent study utilizing the rat optic system demonstrated that [14C-acetyl]-NAA originating in retinal ganglion neurons provided acetyl for lipid synthesis within myelin of the optic system (Chakraborty et al., 2001). This suggested transaxonal movement of NAA from neuron to myelin, following by hydrolysis via myelin-localized ASPA. The presence of ASPA in myelin provides an interesting parallel to the numerous other lipid-synthesizing and metabolizing enzymes that have been detected as integral components of purified myelin (for review: Norton and Cammer, 1984; Ledeen, 1992). Especially significant in the present context are the fatty acid synthesizing enzymes, acetyl-CoA carboxylase and fatty acid synthase (Chakraborty and Ledeen, 2003); the latter enzyme complex in purified myelin showed a level of activity approximately half that of its cytosolic counterpart, and differed from the latter in requiring detergent. These are the enzymes that would utilize ASPA-liberated acetyl groups to form fatty acids, which then become incorporated into myelin phospholipids. Several myelin intrinsic enzymes that catalyze the latter reactions have been identified, including those that mediate the Kennedy pathway for synthesis of choline- and ethanolamine-phosphoglycerides (Ledeen, 1992). These findings indicate NAA can be added to the list of precursors shown to undergo axon-to-myelin transfer with subsequent incorporation into myelin lipids. These include choline (Droz et al., 1978; 1981), phosphate (Ledeen and Haley, 1983), acyl chains (Toews and Morell, 1981; Alberghina et al., 1982), and serine (Haley and Ledeen, 1979). Yet to be elucidated is the mechanism by which such axon-to-myelin transfer proceeds. Figure 5 depicts the proposed pathways for NAA metabolism, transfer, and utilization within myelin.

This study has provided additional correlative evidence for a role of NAA in myelinogenesis in comparing rates of NAA synthesis (ANAT) in various neuronal types. This was seen most clearly in the robust rate of NAA synthesis in adult pig retinal sections containing retinal ganglion neurons, the one retinal neuron type whose axon is myelinated; ANAT rate was ~8-fold greater in those sections than in retinal sections enriched in photoreceptor cells whose axons are not myelinated (Fig. 3A). A lesser though still significant difference was seen in ANAT activity in CN vs CGN (Fig. 3B), the former containing some neurons destined to extend myelinated axons in contrast to CGN that do not. That the difference was not greater was likely due to the mixed neuronal types in CN and the early developmental stage. Complete absence of ANAT in neurons with non-myelinated axons would not be expected since such cells might well require ANAT for synthesis of NAAG, an abundant neurotransmitter (Gehl et al., 2004). This study further highlights the bimodal distribution of ANAT in microsomes and mitochondria, reported in our recent study (Lu et al., 2004). These findings suggest the possibility of compartmentalization, microsomal ANAT producing NAA for myelin and the mitochondrial product perhaps being utilized for NAAG synthesis. However, there is no supporting evidence as yet for this speculative idea.

Figure 5. Cartoon depiction of our working hypothesis showing neuronal synthesis of NAA (by ANAT) in both microsomes and mitochondria, followed by transaxonal transfer to myelin and release of acetyl groups within myelin by aspartoacylase (ASPA). Myelin-associated enzymes, such as fatty acid synthase (FAS), utilize the liberated acetyl groups for fatty acid synthesis followed by incorporation into myelin lipids by other myelin-localized enzymes. Liberated aspartate (Asp) is recycled back to the neuron. Absence of ASPA is postulated to block this pathway of myelin lipid synthesis.

Figure 5. Cartoon depiction of our working hypothesis showing neuronal synthesis of NAA (by ANAT) in both microsomes and mitochondria, followed by transaxonal transfer to myelin and release of acetyl groups within myelin by aspartoacylase (ASPA). Myelin-associated enzymes, such as fatty acid synthase (FAS), utilize the liberated acetyl groups for fatty acid synthesis followed by incorporation into myelin lipids by other myelin-localized enzymes. Liberated aspartate (Asp) is recycled back to the neuron. Absence of ASPA is postulated to block this pathway of myelin lipid synthesis.

In proposing this role for NAA in relation to myelinogenesis it is appropriate to emphasize that this is not conceived as the sole or even major pathway for myelin lipid synthesis. Myelination proceeds to a significant degree in the absence of ASPA-generated, NAA-derived acetyl groups, obviously through parallel pathways residing within the OL. Nevertheless inability to utilize NAA-acetyl leads to significant lipid deficits within myelin that may contribute to the observed pathology and nervous system dysfunction in Canavan disease. Analysis of other myelin lipids from ASPA-null animals may further elucidate this deficit as it affects myelin not only of Canavan patients but also victims of other neurological disorders characterized by subnormal levels of NAA in brain.

Was this article helpful?

0 0

Post a comment